A few years ago, back around Iris v1.0, I developed a bare-bones loader for 'virtual datasets' encoded using the NetCDF Markup Language (
NcML). It wasn't especially robust or feature-complete, so I put it on the back-burner.
However, the recent enhancements to Iris's concatenation and merge capabilities mean that implementing an NcML file loader is - or at least should be - a somewhat more tractable task. Accordingly I've reworked my original code into something that is hopefully a little bit more useful and robust, with the caveat that it very definitely falls into the
experimental category, and I suspect it probably won't handle large and/or complicated NcML files particularly well.
Although Iris in its current form provides support for a number of NcML-like aggregation operations, you may still find some of the extra NcML features useful.
Anyway, if you're interested in having a play with this experimental NcML file loader hop over to my
github page, where you'll find information regarding the Iris branch that you'll need to grab, plus some basic NcML usage examples.
If people find it useful (and usable) then I'll endeavour to extend and refine it. Constructive feedback welcome.
Regards,
Phil